A binding robot system and a path planning method and a work method thereof

Abstract

The application provides a binding robot system and a path planning method and a working method thereof, the path planning method comprising: global planning: performing rapid and low-dimensional rough planning in a task space, establishing a safe buffer zone around an obstacle by introducing a concept of an extended distance; adopting a bidirectional search strategy to improve search efficiency; and generating a safe and smooth 'path main line' through a smoothing optimization strategy; local planning: performing real-time local planning under the guidance of the global 'path main line' by using a neural network algorithm based on a QR-DQN architecture to obtain an optimal binding path; and the neural network algorithm is trained by integrating target guidance and obstacle avoidance as heuristic knowledge into a reward function. Compared with the prior art, the binding robot system and the path planning method and the working method thereof provided by the application can significantly improve the adaptability, efficiency and reliability of robot binding.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing technology, and in particular to a binding robot system and its path planning and operation methods. Background Technology

[0002] Currently, while tying robots can perform simple tasks such as small-scale, low-density, and single-work-face operations, their application efficiency is significantly limited when facing complex scenarios such as large-scale, high-density, and multi-work-face nodes. This is manifested in systemic problems such as frequent spatial conflicts, blind spots in perception, poor reachability of the robotic arm, and complex path planning. The specific bottlenecks can be summarized into the following four categories: ① Limited operating range: Due to the physical arm span of the robotic arm, existing systems can usually only cover a single working surface, making it difficult to achieve complete binding of the entire working plane of large steel cages; ② Ground space occupation: The robot body is deployed on the ground, which not only occupies the already tight working space, but also easily interferes with other equipment and personnel on site, affecting the overall construction flow; ③ Insufficient structural accessibility: Traditional robot configurations are difficult to penetrate complex steel cages with narrow internal spaces, such as U-shaped and box-shaped cages, resulting in the inability to guarantee the quality of internal node binding; ④ Inefficient planning algorithms: In complex, high-dimensional tethered environments, traditional path planning algorithms are inefficient, have a high risk of collisions, and struggle to achieve the optimal balance between path smoothness and motion energy consumption.

[0003] Therefore, there is an urgent need for an automated binding solution that has a large working range, high flexibility, can adapt to complex spatial structures, and is intelligent and efficient. Summary of the Invention

[0004] To address the technical problems in the prior art, this invention provides a tying robot system and its path planning and operation methods that can improve the adaptability, efficiency and reliability of tying.

[0005] To achieve the above-mentioned objectives, the present invention provides the following technical solution: A path planning method for a tying robot system, the tying robot system comprising a nine-axis motion module, a machine vision module, and a central control module, wherein the nine-axis motion module constitutes a nine-degree-of-freedom redundant motion system; the machine vision module is used to perform three-dimensional scanning of the rebar cage, acquire high-precision point cloud data, identify rebar intersection nodes, and construct and update an environmental map in real time; the central control module is used to receive node information and the environmental map from the vision module, plan the tying node sequence, and perform path planning for each segment of the sequence; the path planning method includes: Global planning: Perform rapid, low-dimensional coarse planning in the task space, establish a safety buffer zone around obstacles by introducing the concept of "extended distance"; adopt a bidirectional search strategy to improve search efficiency; and generate a safe and smooth "path mainline" through a smoothing optimization strategy. Local planning: Real-time local planning is performed using a neural network algorithm based on the QR-DQN architecture under the guidance of the global "path mainline" to obtain the optimal binding path; the neural network algorithm is formed by integrating target orientation and obstacle avoidance as heuristic knowledge into the reward function and training the neural network.

[0006] Preferably, the global planning adopts the SSB-A* algorithm, which includes three steps: extended distance optimization, bidirectional search, and smoothing processing.

[0007] Preferably, the expansion distance optimization step includes: expanding obstacles outward in units of grids, the value of which is determined by the robot speed, size and number of grids; the expanded nodes are not visited during the search process.

[0008] Preferably, the bidirectional search step includes: simultaneously searching from the starting node and the ending node; when the forward and reverse search nodes become adjacent nodes, the search process ends, and the two incomplete paths are joined to form a complete collision-free path.

[0009] Preferably, the smoothing step is used to smooth and optimize the path formed by the bidirectional search step, including: When there are no obstacles inside the corner to reduce the turning angle, the inflection point and its two adjacent nodes are replaced with the adjacent points in the corner, and the 90° right-angle turn is decomposed into two 45° acute-angle turns. Or / and, when there are consecutive right-angle turns, the inflection point can be removed, and the two adjacent nodes of the inflection point are directly connected, converting multiple right-angle turns into a small number of 45° acute-angle turns.

[0010] Preferably, in the local planning step, neural network training includes: A local path planning environment model and intelligent agent are established based on the value distribution algorithm; The local path planning problem of a robot is formulated as a Markov decision process problem; The agent perceives environmental information by processing data from the machine vision module and selects either a completely random action or the optimal action under the current neural network policy based on a decaying epsilon-greedy policy. The current reward is calculated based on the heuristic reward function, the tuple is stored in the experience pool, and a batch of experiences is randomly selected from the experience pool to update the neural network parameters. The heuristic reward function is formed by integrating goal orientation and obstacle avoidance as heuristic knowledge into the reward function. The evaluation process considers whether the policy converges to the optimal policy. If it does not converge, the environment is reset to start a new round and training continues. If convergence is achieved, training is terminated, and the trained neural network is obtained.

[0011] A tying robot system, comprising: A nine-axis motion module is used to construct a redundant motion system with nine degrees of freedom, including a robotic arm; The machine vision module is used to perform 3D scanning of the rebar cage, acquire high-precision point cloud data, identify rebar intersections, and build and update the environmental map in real time. The binding execution module, located on the robotic arm, is used to perform rebar binding operations; The central control module connects the nine-axis motion module, the machine vision module, and the tying execution module. It receives node information and environmental maps from the vision module, performs hierarchical path planning, generates the optimal tying path, and controls the nine-axis motion module and the tying execution module to work together to complete the task according to the optimal tying path. The hierarchical path planning adopts the path planning method of the tying robot system described above.

[0012] Preferably, the nine-axis motion module further includes a gantry and a vertical moving device. The gantry spans across the working area of ​​the steel cage and can move along the X-axis on the ground rails laid on both sides of the gantry. The robotic arm is a six-axis articulated robotic arm, connected to the crossbeam via the vertical moving device, so that the robotic arm hangs upside down on the gantry. The vertical moving device can move along the crossbeam to drive the robotic arm to move along the Y-axis of the gantry. The vertical moving device is used to drive the robotic arm to move vertically, so that the robotic arm moves along the Z-axis of the gantry. This allows the robotic arm to move along the X, Y, and Z axes of the gantry, and combined with the six degrees of freedom of the robotic arm itself, it constitutes a nine-degree-of-freedom redundant motion system.

[0013] Preferably, before hierarchical path planning, the central control module also performs sequence optimization, which includes: planning the tying node sequence using an improved traveling salesman problem algorithm, with the goal of minimizing the total moving distance of the robotic arm and the number of area switching.

[0014] A method for operating the tying robot system as described above includes: The machine vision module performs 3D scanning of the rebar cage to obtain high-precision point cloud data, identify rebar intersections, and build and update the environmental map in real time. The central control module receives node information and environmental map from the vision module, performs hierarchical path planning, and generates the optimal binding path. The central control module controls the movement of the nine-axis motion module according to the optimal binding path, so that the nine-axis motion module moves to the defined working area and the robotic arm moves to the binding point; The binding execution module performs the binding operation.

[0015] Compared with existing technologies, the tying robot system, path planning method, and operation method provided by this invention can significantly improve the adaptability, efficiency, and reliability of robot tying by constructing a nine-degree-of-freedom redundant motion system and combining it with the ARPPS path planning algorithm with hierarchical planning ideas. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a three-dimensional structural diagram of a tying robot system provided in one embodiment of the present invention; Figure 2 for Figure 1 A three-dimensional structural diagram of the Y-axis movement device in the tying robot system shown. Figure 3 for Figure 1 Top view of the tying robot system shown; Figure 4 for Figure 1 A flowchart illustrating the operation method of the tying robot system shown; Figure 5 for Figure 4 A schematic diagram comparing the global planning extended distance optimization steps in the path planning method of the illustrated task method with the traditional method; Figure 6 for Figure 4 A schematic diagram comparing the global planning bidirectional search steps in the path planning method of the illustrated task method with the traditional method; Figure 7 for Figure 4 A schematic diagram comparing the first strategy of global planning smooth optimization step in the path planning method of the operation method shown with the traditional method; Figure 8 for Figure 4 A schematic diagram comparing the first strategy of global planning smooth optimization step in the path planning method of the operation method shown with the traditional method; Figure 9 for Figure 4 The flowchart illustrates the neural network training process in the local planning step of the path planning method shown in the task method. Detailed Implementation

[0018] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] It should be noted that when a component is referred to as being "fixed to" or "set on" another component, it can be directly on or indirectly set on the other component; when a component is referred to as being "connected to" another component, it can be directly connected to or indirectly connected to the other component.

[0020] It should be noted that the structures, proportions, sizes, etc., shown in the accompanying drawings of this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the conditions under which this application can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size should still fall within the scope of the technical content disclosed in this application, provided that they do not affect the effects and purposes that this application can produce.

[0021] like Figures 1 to 3 As shown, an embodiment of the present invention provides a tying robot system, comprising: A nine-axis motion module is used to construct a redundant motion system with nine degrees of freedom, including a robotic arm 1; The machine vision module is used to perform 3D scanning of the rebar cage 2, acquire high-precision point cloud data, identify rebar intersections, and build and update the environmental map in real time. The binding execution module, located on robotic arm 1, is used to perform rebar binding operations; The central control module, connecting the nine-axis motion module, machine vision module, and tying execution module, receives node information and environmental maps from the vision module, performs hierarchical path planning, generates the optimal tying path, and controls the nine-axis motion module and tying execution module to collaboratively complete the task according to the optimal tying path. In this embodiment, before hierarchical path planning, the central control module also performs sequence optimization, which includes planning the tying node sequence using an improved traveling salesman problem algorithm, with the goal of minimizing the total movement distance of the robotic arm and the number of area switching times.

[0022] In this embodiment, the nine-axis motion module also includes a gantry frame 3 and a vertical moving device 4. The gantry frame 3 spans above the working area of ​​the steel cage and can move along the X-axis on the ground rails 7 laid on both sides of the gantry frame via an X-axis moving device (such as a motor driving the bottom wheels, not shown). A Y-axis moving device is also installed on the crossbeam of the gantry frame 3.

[0023] In this embodiment, the robotic arm 1 is a six-axis articulated robotic arm, connected to the Y-axis moving device 5 on the crossbeam via a vertical moving device 4. The vertical moving device 4 can move along the crossbeam via the Y-axis moving device 5, allowing the robotic arm to hang upside down on the gantry and move along the Y-axis of the gantry. Specifically, in this embodiment, the Y-axis moving device 5 can be a moving conveyor belt structure. One end of the vertical moving device 4 is mounted on the Y-axis moving device 5, and the other end is connected to the robotic arm 1 via a robotic arm connecting end 6, driving the robotic arm to move vertically, thus moving the robotic arm along the Z-axis of the gantry.

[0024] The aforementioned X-axis moving device that moves the gantry along the X-axis, the Y-axis moving device 5 that moves the robotic arm along the Y-axis of the gantry, and the vertical moving device 4 that moves the robotic arm along the Z-axis of the gantry together constitute the XYZ-axis motion mechanism. Through this mechanism, the robotic arm can move along the X and Y axes of the gantry, achieving a very wide range of movement in the horizontal plane. Combined with the robotic arm's own six degrees of freedom motion, this forms a nine-degree-of-freedom redundant motion system, allowing the robotic arm's end effector to reach any position within the workspace in any posture. In this embodiment, the binding robot system uses the XYZ-axis motion mechanism to quickly transport the robotic arm to any macroscopic work unit above the rebar cage. The six-axis articulated robotic arm then performs precise, full-posture binding operations within that unit, enabling the robot to easily cover large rebar cages tens of meters long and flexibly adjust its posture to work inside complex structures such as U-shaped and box-shaped structures—capabilities that traditional structures cannot achieve.

[0025] In this embodiment, the machine vision module uses a structured light camera, which is fixedly installed at the end of the robotic arm or on the gantry beam. It is used to perform three-dimensional scanning of the rebar cage, obtain high-precision point cloud data, identify rebar intersection nodes, and build and update the environmental map in real time.

[0026] In this embodiment, the binding execution module is a binding gun installed at the end of the robotic arm. Its binding wire spool and battery adopt a distributed design and are integrated inside the robotic arm's upper arm or gantry. The wire is fed through the built-in channel, which greatly reduces the load at the end and avoids cable dragging.

[0027] The tying robot system of this invention adopts an inverted configuration, which has the following significant advantages: ① Enhance perception field of view and reduce blind spots: The high installation position provides the visual sensor with a wider global overview view, which helps to perceive the distribution and status of the rebar nodes more comprehensively and significantly reduces the perception blind spots caused by obstruction. ② Optimize space utilization and avoid ground interference: The inverted installation allows the robot body to be removed from the ground working area, which can effectively avoid the obstruction caused by the dense steel mesh to the robot's movement, and fundamentally reduce the spatial conflict with complex ground environment and existing steel structure. ③ Enhanced end-effector accessibility: The inverted configuration allows the robotic arm to approach the work point in a "top-down" manner, which is particularly suitable for vertical binding operations in high-density steel mesh, improving the end effector's accessibility and operational flexibility to multiple work surface nodes. ④ Supports modular and scalable deployment: The inverted robot can be deployed using a track or gantry system, making it easy to achieve long-distance movement across work surfaces and multi-robot collaborative operation, which provides convenience for system expansion in large-scale construction scenarios.

[0028] The aforementioned hierarchical path planning is a path planning method for a tethering robot system provided in this embodiment (also known as ARPPS, Autonomous Robotic Path Planning System). Figure 4 As shown, this path planning method includes two main steps: global planning and local planning.

[0029] In global planning, rapid, low-dimensional coarse planning is performed in the task space. By introducing the concept of "extended distance", a safe buffer zone is established around obstacles. A bidirectional search strategy is adopted to improve search efficiency. A smooth optimization strategy is used to generate a safe and smooth "path mainline".

[0030] Specifically, in this embodiment, the global planning adopts the SSB-A* (Safe, Smooth, and BidirectionalA*) algorithm, which includes three steps: extended distance optimization, bidirectional search, and smoothing processing.

[0031] The SSB-A* algorithm introduces the concept of extended distance on top of the traditional A* algorithm: it preserves an extra safety space around obstacles during path planning (e.g., ...). Figure 5 As shown, the left image represents the traditional algorithm, and the right image represents the proposed algorithm (which expands the middle gray space). Specifically, the expansion distance optimization steps include: expanding obstacles outward in grid units, the value of which is determined by the robot's speed, size, and number of grids; the expanded nodes are not visited during the search process, serving both as collision buffers to reduce risk and improving algorithm efficiency by reducing the effective graph size.

[0032] The bidirectional search steps include: simultaneously searching from both the starting and ending nodes; the search process ends when nodes searched in opposite directions become adjacent, and the two incomplete paths are joined to form a complete, collision-free path. The traditional A* algorithm's unidirectional search mode (such as...) Figure 6 (As shown in the middle left figure) has limited efficiency. This algorithm adopts a bidirectional search method (such as...) Figure 6 (As shown in the middle right figure) can significantly improve search efficiency.

[0033] The smoothing step is used to optimize the path formed by the bidirectional search step. The traditional A* algorithm path consists of nodes and polylines, and has three main drawbacks: it does not minimize the number of vertices, the path is discontinuous, and right-angle turns negatively impact motion performance. This algorithm proposes two smoothing optimization strategies specifically for the right-angle turn problem: When there are no obstacles inside the corner to reduce the turning angle, the inflection point and its two adjacent nodes are replaced with adjacent points in the corner, and a 90° right-angle turn is made (e.g., Figure 7 As shown in the middle left figure, the path of the traditional algorithm is decomposed into two 45° acute-angle turns (such as...). Figure 7 The middle right figure shows the path optimized by this algorithm. When there are consecutive right-angle turns, the inflection point can be removed, and the two adjacent nodes of the inflection point are directly connected, thus removing multiple right-angle turns (such as...). Figure 8 As shown in the middle left figure, the path of the traditional algorithm is converted into a small number of 45° acute angle turns (such as...). Figure 8 The right-hand side of the figure shows the path optimized by this algorithm.

[0034] like Figure 4 As shown, the local planning adopts the ARLOA (Adaptive Reinforcement Learning ObstacleAvoidance) algorithm, which mainly includes: using a neural network algorithm based on the QR-DQN (a reinforcement learning algorithm based on the ideas of deep Q-learning and quantile regression) architecture to perform real-time local planning under the guidance of the global "path mainline" to obtain the optimal binding path; the neural network algorithm is formed by integrating target orientation and obstacle avoidance as heuristic knowledge into the reward function and training the neural network.

[0035] like Figure 9 As shown, in this embodiment, neural network training includes: A local path planning environment model and intelligent agent are established based on the value distribution algorithm; The local path planning problem of a robot is formulated as a Markov decision process problem; The agent perceives environmental information by processing data from the machine vision module and selects either a completely random action or the optimal action under the current neural network policy based on a decaying epsilon-greedy policy. The current reward is calculated based on the heuristic reward function, the tuple is stored in the experience pool, and a batch of experiences is randomly selected from the experience pool to update the neural network parameters. The heuristic reward function is formed by integrating goal orientation and obstacle avoidance as heuristic knowledge into the reward function. The evaluation process considers whether the policy converges to the optimal policy. If it does not converge, the environment is reset to start a new round and training continues. If convergence is achieved, training is terminated, and the trained neural network is obtained.

[0036] The path planning method in this embodiment can also perform execution and closed-loop control. That is, the robotic arm executes the planned trajectory, during which the visual servo system performs real-time pose fine-tuning, and the force sensor monitors the binding quality. The path planning method is adjusted and optimized based on the adjusted pose and the monitored binding quality. Furthermore, if a significant change occurs in the environment during path planning, the algorithm is triggered to perform local map updates and path replanning.

[0037] The path planning method in this embodiment significantly reduces computational complexity and generates high-quality, executable optimal trajectories. It can also achieve intelligent obstacle avoidance and motion optimization through "regional weight switching" (prioritizing safety in narrow passages and efficiency in open areas).

[0038] Specifically, traditional path planning algorithms (such as A* and RRT) perform well in solving general point-to-point navigation problems for mobile robots, but they fall short when applied to complex, high-dimensional, multi-constraint, and dynamic scenarios like rebar tying. The ARPPS algorithm in this embodiment is designed to address these challenges, offering advantages that are systematic and multi-layered.

[0039]

[0040] This embodiment also provides a method for operating the aforementioned tying robot system, including: The machine vision module performs 3D scanning of the rebar cage to obtain high-precision point cloud data, identify rebar intersections, and build and update the environmental map in real time. The central control module receives node information and environmental map from the vision module, performs hierarchical path planning, and generates the optimal binding path. The central control module controls the movement of the nine-axis motion module according to the optimal binding path, so that the nine-axis motion module moves to the defined working area and the robotic arm moves to the binding point; The binding execution module performs the binding operation.

[0041] During movement, the nine-axis motion module uses XYZ axis motion mechanisms for large-scale position adjustments, resulting in higher efficiency; while the six-axis robotic arm handles precise attitude adjustments, ensuring the tying gun is accurately aligned with the node. Upon reaching the node, the tying gun mounted at the end of the robotic arm performs the tying operation.

[0042] The above description is merely an embodiment of the present invention. It should be noted that those skilled in the art can make improvements without departing from the inventive concept of the present invention, but these improvements all fall within the protection scope of the present invention.

Claims

1. A path planning method for a tying robot system, characterized in that, The binding robot system includes a nine-axis motion module, a machine vision module, and a central control module. The nine-axis motion module constitutes a nine-degree-of-freedom redundant motion system. The machine vision module is used to perform three-dimensional scanning of the rebar cage, acquire high-precision point cloud data, identify rebar intersection nodes, and construct and update the environmental map in real time. The central control module is used to receive node information and the environmental map from the vision module, plan the binding node sequence, and perform path planning for each path segment in the sequence. The path planning method includes: Global planning: Perform rapid, low-dimensional coarse planning in the task space, establish a safety buffer zone around obstacles by introducing the concept of "extended distance"; adopt a bidirectional search strategy to improve search efficiency; and generate a safe and smooth "path mainline" through a smooth optimization strategy. Local planning: Real-time local planning is performed using a neural network algorithm based on the QR-DQN architecture under the guidance of the global "path mainline" to obtain the optimal binding path; the neural network algorithm is formed by integrating target orientation and obstacle avoidance as heuristic knowledge into the reward function and training the neural network.

2. The path planning method for the tying robot system according to claim 1, characterized in that, The global planning uses the SSB-A* algorithm, which includes three steps: extended distance optimization, bidirectional search, and smoothing.

3. The path planning method for the tying robot system according to claim 2, characterized in that, The expansion distance optimization step includes: expanding obstacles outward in units of grids, the value of which is determined by the robot's speed, size, and number of grids; the expanded nodes are not visited during the search process.

4. The path planning method for the tying robot system according to claim 3, characterized in that, The bidirectional search steps include: simultaneously searching from the starting node and the ending node; when the forward and reverse search nodes become adjacent nodes, the search process ends, and the two incomplete paths are joined together to form a complete collision-free path.

5. The path planning method for the tying robot system according to claim 4, characterized in that, The smoothing process step is used to smooth and optimize the path formed by the bidirectional search step, including: When there are no obstacles inside the corner to reduce the turning angle, the inflection point and its two adjacent nodes are replaced with the adjacent points in the corner, and the 90° right-angle turn is decomposed into two 45° acute-angle turns. Or / and, when there are consecutive right-angle turns, the inflection point can be removed, and the two adjacent nodes of the inflection point are directly connected, converting multiple right-angle turns into a small number of 45° acute-angle turns.

6. The path planning method for the tying robot system according to claim 1, characterized in that, In the local planning step, neural network training includes: A local path planning environment model and intelligent agent are established based on the value distribution algorithm; The local path planning problem of a robot is formulated as a Markov decision process problem; The agent perceives environmental information by processing data from the machine vision module and selects either a completely random action or the optimal action under the current neural network policy based on a decaying epsilon-greedy policy. The current reward is calculated based on the heuristic reward function, the tuple is stored in the experience pool, and a batch of experiences is randomly selected from the experience pool to update the neural network parameters. The heuristic reward function is formed by integrating goal orientation and obstacle avoidance as heuristic knowledge into the reward function. The evaluation process considers whether the policy converges to the optimal policy. If it does not converge, the environment is reset to start a new round and training continues. If convergence is achieved, training is terminated, and the trained neural network is obtained.

7. A tying robot system, characterized in that, include: A nine-axis motion module is used to construct a redundant motion system with nine degrees of freedom, including a robotic arm; The machine vision module is used to perform 3D scanning of the rebar cage, acquire high-precision point cloud data, identify rebar intersections, and build and update the environmental map in real time. The binding execution module, located on the robotic arm, is used to perform rebar binding operations; The central control module, connected to the nine-axis motion module, machine vision module, and tying execution module, is used to receive node information and environmental map from the vision module, perform hierarchical path planning, generate the optimal tying path, and control the nine-axis motion module and tying execution module to work together to complete the task according to the optimal tying path; the hierarchical path planning adopts the path planning method of the tying robot system as described in any one of claims 1 to 6.

8. The tying robot system according to claim 7, characterized in that, The nine-axis motion module also includes a gantry and a vertical moving device. The gantry spans above the working area of ​​the steel cage and can move along the X-axis on the ground rails laid on both sides of the gantry. The robotic arm is a six-axis articulated robotic arm, connected to the crossbeam via the vertical moving device, so that the robotic arm hangs upside down on the gantry. The vertical moving device can move along the crossbeam to drive the robotic arm to move along the Y-axis of the gantry. The vertical moving device is used to drive the robotic arm to move vertically, so that the robotic arm moves along the Z-axis of the gantry. This allows the robotic arm to move along the X, Y, and Z axes of the gantry, which, combined with the six degrees of freedom of the robotic arm itself, constitutes a nine-degree-of-freedom redundant motion system.

9. The tying robot system according to claim 7, characterized in that, Before hierarchical path planning, the central control module also performs sequence optimization, which includes: planning the binding node sequence using an improved traveling salesman problem algorithm, with the goal of minimizing the total moving distance of the robotic arm and the number of area switching.

10. A method for operating the tying robot system as described in claim 7, 8, or 9, characterized in that, include: The machine vision module performs 3D scanning of the rebar cage to obtain high-precision point cloud data, identify rebar intersections, and build and update the environmental map in real time. The central control module receives node information and environmental map from the vision module, performs hierarchical path planning, and generates the optimal binding path. The central control module controls the movement of the nine-axis motion module according to the optimal binding path, so that the nine-axis motion module moves to the defined working area and the robotic arm moves to the binding point; The binding execution module performs the binding operation.

Images